NEC

About: NEC is a company organization based out in Tokyo, Japan. It is known for research contribution in the topics: Signal & Layer (electronics). The organization has 33269 authors who have published 57670 publications receiving 835952 citations. The organization is also known as: NEC Corporation & NEC Electronics Corporation.

Simplifying Graph Convolutional Networks

Abstract: Graph Convolutional Networks (GCNs) and their variants have experienced significant attention and have become the de facto methods for learning graph representations. GCNs derive inspiration primarily from recent deep learning approaches, and as a result, may inherit unnecessary complexity and redundant computation. In this paper, we reduce this excess complexity through successively removing nonlinearities and collapsing weight matrices between consecutive layers. We theoretically analyze the resulting linear model and show that it corresponds to a fixed low-pass filter followed by a linear classifier. Notably, our experimental evaluation demonstrates that these simplifications do not negatively impact accuracy in many downstream applications. Moreover, the resulting model scales to larger datasets, is naturally interpretable, and yields up to two orders of magnitude speedup over FastGCN.

Capillarity-induced filling of carbon nanotubes

Abstract: THE recent discovery1 and bulk synthesis2 of nanometre-scale carbon tubules has led to much speculation about possible uses of these graphitic structures3–5. Broughton and Pederson predicted on the basis of computer simulations that open nanotubes may be filled with liquid by capillary suction6. Here we describe experiments in which annealing of the tubules in the presence of liquid lead results in opening of the capped tube ends and subsequent filling of the tubes with molten material through capillary action. The nanotubes thus act as moulds for the fabrication of (possibly metallic) wires, some of which are less than two nanometres in diameter.

FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

Abstract: Semi-supervised learning (SSL) provides an effective means of leveraging unlabeled data to improve a model's performance. In this paper, we demonstrate the power of a simple combination of two common SSL methods: consistency regularization and pseudo-labeling. Our algorithm, FixMatch, first generates pseudo-labels using the model's predictions on weakly-augmented unlabeled images. For a given image, the pseudo-label is only retained if the model produces a high-confidence prediction. The model is then trained to predict the pseudo-label when fed a strongly-augmented version of the same image. Despite its simplicity, we show that FixMatch achieves state-of-the-art performance across a variety of standard semi-supervised learning benchmarks, including 94.93% accuracy on CIFAR-10 with 250 labels and 88.61% accuracy with 40 -- just 4 labels per class. Since FixMatch bears many similarities to existing SSL methods that achieve worse performance, we carry out an extensive ablation study to tease apart the experimental factors that are most important to FixMatch's success. We make our code available at this https URL.

ONOS: towards an open, distributed SDN OS

Abstract: We present our experiences to date building ONOS (Open Network Operating System), an experimental distributed SDN control platform motivated by the performance, scalability, and availability requirements of large operator networks. We describe and evaluate two ONOS prototypes. The first version implemented core features: a distributed, but logically centralized, global network view; scale-out; and fault tolerance. The second version focused on improving performance. Based on experience with these prototypes, we identify additional steps that will be required for ONOS to support use cases such as core network traffic engineering and scheduling, and to become a usable open source, distributed network OS platform that the SDN community can build upon.

Opening carbon nanotubes with oxygen and implications for filling

Abstract: CAPPED hollow carbon nanotubes1,2 can be modified into nanocomposite fibres by simultaneous opening of the caps (by heating in the presence of air and lead metal) and filling of the interior with an inorganic phase3. To generalize this approach, greater understanding is needed of the reaction mechanism between the tube caps and oxygen. Here we report that the oxidation of carbon nanotubes in air for short durations above about 700 °C results in the etching away of the tube caps and the thinning of tubes through layer-by-layer peeling of the outer layers, starting from the cap region. The oxidation reaction follows an Arrhenius-type relation with an activation energy barrier of about 225 kJ mol−1 in air. Heating of closed nanotubes with an oxide (Pb3O4) in an inert atmosphere lowers the activation barrier for the reaction and opening of the tubes occurs at lower temperatures. Contrary to intuition, however, open tubes are much more difficult to fill with inorganic materials than in the one-step filling of tubes reported previously3. But various other experiments might be possible in the inner nano-cavities of the open tubes such as studies of catalysis and of low-dimensional chemistry and physics.